def r_faces(self):
if self.image is not None:
faceCascade = cv2.CascadeClassifier('cascades/face.xml')
eyeCascade = cv2.CascadeClassifier('cascades/haarcascade_eye.xml')
faceSize = config.DEFAULT_FACE_SIZE
threshold = 500
recognizer = train.trainRecognizer('train',
faceSize,
showFaces=True)
capture = cv2.imread(self.img)
while True:
self.image = imutils.resize(capture, height=500)
for (label, confidence,
(x, y, w,
h)) in reconocer.RecognizeFace(self.image, faceCascade,
eyeCascade, faceSize,
threshold):
cv2.rectangle(self.image, (x, y), (x + w, y + h),
(0, 255, 0), 2)
cv2.putText(
self.image,
"{} = {}".format(recognizer.getLabelInfo(label),
int(confidence)), (x, y),
cv2.FONT_HERSHEY_PLAIN, 1, (0, 255, 0), 1, cv2.LINE_AA)
self.mostrarImagen()
break
def RecognizeFace(image, faceCascade, eyeCascade, faceSize, threshold):
found_faces = []
recognizer = train.trainRecognizer("train", faceSize, showFaces=True)
gray, faces = detect.detectFaces(image, faceCascade, eyeCascade, returnGray=1)
for ((x, y, w, h), eyedim) in faces:
label, confidence = recognizer.predict(cv2.resize(detect.levelFace(gray, ((x, y, w, h), eyedim)), faceSize))
if confidence < threshold:
found_faces.append((label, confidence, (x, y, w, h)))
return found_faces
def prepareImage(filename, FEDetector, cascadeDetector):
haarFaceCascade = cv2.CascadeClassifier(config.HAAR_FACE_CASCADE_FILE)
lbpFaceCascade = cv2.CascadeClassifier(config.LBP_FACE_CASCADE_FILE)
haarEyeCascade = cv2.CascadeClassifier(config.HAAR_EYE_CASCADE_FILE)
faceSize = config.DEFAULT_FACE_SIZE
threshold = 500
recognizer = train.trainRecognizer('Train Set',
FEDetector,
cascadeDetector,
faceSize,
showFaces=True)
print "recognizer"
print cascadeDetector
print FEDetector
# cv2.namedWindow("camera", 1)
# capture = cv2.VideoCapture(0)
# path = "/Users/Mohamad/AUT/B.Sc. Thesis/SR_FR Project/Resources/Test Set/image_0022.jpg"
path = filename
# while True:
# retval, img = capture.read()
haar_img = cv2.imread(path)
lbp_img = cv2.imread(path)
if cascadeDetector == "1":
# HAAR CASCADE
for (label, confidence,
(x, y, w, h)) in recognizeFace(haar_img, haarFaceCascade,
haarEyeCascade, faceSize,
threshold, recognizer):
cv2.rectangle(haar_img, (x, y), (x + w, y + h), (0, 255, 0), 2)
cv2.putText(
haar_img, "{} = {}".format(recognizer.getLabelInfo(label),
int(confidence)), (x, y),
cv2.FONT_HERSHEY_PLAIN, 1, (0, 255, 0), 1, cv2.LINE_AA)
cv2.imwrite("Resources/Recognized/haar_recognized.jpg", haar_img)
recognizedPath = "/Users/Mohamad/AUT/B.Sc. Thesis/SR_FR Project/Resources/Recognized/haar_recognized.jpg"
return recognizedPath
elif cascadeDetector == "2":
# LBP CASCADE
for (label, confidence,
(x, y, w, h)) in recognizeFace(lbp_img, lbpFaceCascade, None,
faceSize, threshold, recognizer):
cv2.rectangle(lbp_img, (x, y), (x + w, y + h), (0, 255, 0), 2)
cv2.putText(
lbp_img, "{} = {}".format(recognizer.getLabelInfo(label),
int(confidence)), (x, y),
cv2.FONT_HERSHEY_PLAIN, 1, (0, 255, 0), 1, cv2.LINE_AA)
cv2.imwrite("Resources/Recognized/lbp_recognized.jpg", lbp_img)
recognizedPath = "/Users/Mohamad/AUT/B.Sc. Thesis/SR_FR Project/Resources/Recognized/lbp_recognized.jpg"
return recognizedPath
def reconocer(image):
# Lee argumentos
faceCascade = cv2.CascadeClassifier('cascades/face.xml')
eyeCascade = cv2.CascadeClassifier('cascades/haarcascade_eye.xml')
faceSize = config.DEFAULT_FACE_SIZE
threshold = 500
recognizer = train.trainRecognizer('train', faceSize, showFaces=True)
# Crea la ventana con el nombre 'Reconocimiento Facial!'
# cv2.namedWindow("Reconocimiento Facial!", 1)
# Pasa como parametro la imagen recibida como argumento
array = numpy.frombuffer(image, dtype='uint8')
capture = cv2.imdecode(array, -1)
results = []
for (label, confidence,
(x, y, w, h)) in RecognizeFace(capture, faceCascade, eyeCascade,
faceSize, threshold):
results.append(
dict(label=recognizer.getLabelInfo(label), confidence=confidence))
result = {"results": results}
print(result)
return result
# returns the label only.
#label = recognizer.predict(cv2.resize(detect.levelFace(gray, ((x, y, w, h), eyedim)), faceSize))
#confidence = -1
if confidence < threshold:
found_faces.append((label, confidence, (x, y, w, h)))
return found_faces
if __name__ == '__main__':
faceCascade = cv2.CascadeClassifier(config.FACE_CASCADE_FILE)
eyeCascade = cv2.CascadeClassifier(config.EYE_CASCADE_FILE)
faceSize = config.DEFAULT_FACE_SIZE
threshold = 500
recognizer = train.trainRecognizer('imgdb', faceSize, showFaces=True)
cv2.namedWindow("camera", 1)
capture = cv2.VideoCapture(0)
while True:
retval, img = capture.read()
for (label, confidence, (x, y, w, h)) in RecognizeFace(img, faceCascade, eyeCascade, faceSize, threshold):
cv2.rectangle(img, (x, y), (x+w, y+h), (0, 255, 0), 2)
cv2.putText(img, "{} = {}".format(recognizer.getLabelInfo(label), int(confidence)), (x, y), cv2.FONT_HERSHEY_PLAIN, 1, (0,255,0), 1, cv2.LINE_AA)
cv2.imshow("camera", img)
if cv2.waitKey(30) & 0xFF == ord('q'):
break
for ((x, y, w, h), eyedim) in faces:
label, confidence = recognizer.predict(cv2.resize(detection.levelFace(gray, ((x, y, w, h), eyedim)), faceSize))
if confidence > threshold:
found_faces.append((label, confidence, (x, y, w, h)))
return found_faces
if __name__ == '__main__':
faceCascade = cv2.CascadeClassifier(resource.FACE_CASCADE)
eyeCascade = cv2.CascadeClassifier(resource.EYE_CASCADE)
faceSize = resource.DEFAULT_FACE_SIZE
threshold = 100
recognizer = train.trainRecognizer('imgdb', faceSize)
cv2.namedWindow("camera", 1)
capture = cv2.VideoCapture(0)
while True:
retval, img = capture.read()
for (label, confidence, (x, y, w, h)) in RecognizeFace(img, faceCascade, eyeCascade, faceSize, threshold):
cv2.rectangle(img, (x, y), (x+w, y+h), (0, 255, 0), 2)
cv2.putText(img, "{} = {}".format(recognizer.getLabelInfo(label), int(confidence)), (x, y), cv2.FONT_HERSHEY_PLAIN, 1, (0,255,0), 1, cv2.LINE_AA)
cv2.imshow("camera", img)
if cv2.waitKey(30) & 0xFF == ord('q'):
break
def entrenamiento_service():
trainRecognizer('train', showFaces=False, forceTrain=True)
result = "Entrenamiento: OK"
print(result)
return jsonify(result)